Abstract Growing supermassive black holes ( ) that power luminous quasars from light seeds—the remnants of the first stars—within a Gyr of the Big Bang poses a timing challenge. The formation of massive black hole seeds via direct collapse with initial masses alleviates this problem. Viable direct-collapse black hole formation sites, the satellite halos of star-forming galaxies, merge and acquire stars to produce a new, transient class of high-redshift objects, obese black hole galaxies (OBGs). The accretion luminosity outshines that of the stars in OBGs. We predict the multi-wavelength energy output of OBGs and growing Pop III remnants at z = 9 for standard and slim disk accretion, as well as high and low metallicities of the associated stellar population. We derive robust selection criteria for OBGs—a pre-selection to eliminate blue sources, followed by color–color cuts and the ratio of X-ray flux to rest-frame optical flux . Our cuts sift out OBGs from other bright, high- and low-redshift contaminants in the infrared. OBGs with predicted are unambiguously detectable by the Mid-Infrared Instrument (MIRI), on the upcoming James Webb Space Telescope (JWST). For parameters explored here, growing Pop III remnants with predicted will likely be undetectable by JWST. We demonstrate that JWST has the power to discriminate between initial seeding mechanisms.